Set up using AMC AB15A100 drives, brushed DC motor Prototrak Plus with Encoders

Todd, thanks!! I have several pics of traces that have the trace (top of the corner is not sharp, has radius) not filling (what Pot or SW- fills the corner; what is best way to fill corner?) that corner. Only 1 trace attached, units are uniform, but you see the trace is on bottom of center line, it seems to be more consistence this direction. I have worked through the sw 2 settings to control overshoot, Pot1 as high w/o any noise or vibrations. Setting on pic are y=y axes motor; s2=1.267; s3=SW3 .20 & .0147 P1=11.5 turn cw; P3=10.625 cw fd=20"/min.

Many thanks
 

Who have you spoken with at AMC? I've been speaking with them also about some older analog drives.

Many thanks, Roguish!! Yes, and the engineering team has been excellent. This question of how quickly the motors respond, I give you a quick over view about the original system was "back EMF" style of feedback, I was not as knowledgeable with how the different cnc systems accept feedback at one point in time. The A-M-C AxCent servo drive model AB15A100 has a place for encoder input and now I have encoders that match the drives, with a good wiring diagram I am in the process of tuning the drive to motor with the hope of tuning to the LinuxCnc software very soon. The original transformer (Southwestern Industries) was a 40 volt DC, & as best I can understand this works with a "back emf" style of control. The system was a 1983 propitiate (Prototrack Plus) system, that was well tuned to take full advantage of the "back emf" feedback.

My A-M-C engineer has been absolutely great! I asked which A-M-C power supply is best suited for my application, A-M-C suggested a few different models & I was just crazy lucky to find one on ebay, that looked new.

I am ever so close to getting this mill back to cutting chips with the help and guidance of the forum and its members, I am enormously grateful for all the forum organizers, the moderators, providing a place that brings people and ideas together, with some true wizards, that hang out here.

Many thanks

Moving in the right direction.

What does the trace look like when you set SW2 to off, off, off, on? (leaving every thing else alone.)

Todd, thank you! This attached pics are a close to the trace as the one you requested & a pic of the settings you asked for. I don;t understand why it will not repeat with a very close trace if settings are the same.
These pics are with table moving left to right (vs. right to left) the trace (rt to left) has a great deal of difference, not in the good way. I have spent a great deal of time tuning, just nothing real promising to show. I am staying with the left to rt feed direction.

Many thanks!

In the pics, the only Pot 3 changes, 11.0 vs. 13.5.
 

Does your machine have lead screws or ball screws? The most likely reason for the difference in reaction for different direction moves, is because the friction in one direction might be different from the other. Could be because of a number of reasons. Maybe one of the gibs is a bit out of adjustment and moving in one direction causes the table to turn a little and causes something to bind a bit.

Todd, thanks for your help! This has ball screws. After thinking it could be the sensitivity of the hal scope is on display, because I cannot notice the difference in turning the handles.
I am considering mapping out all the possibilities of SW2 &SW3 (1 to 5 only) and start low with the hope I will find the sweet spot early in the process.

But I am also considering the o-scope procedure. I asked my A-M-C engineer about the speed of the scope the suggestion was 200kHz.

This is a pic of eBay item number: 403709500631, delivered around $30.00, do you think it will work? It claims to do PWM signal output & more. This is also considering using Linuxcnc to generate the signal & using pin 8 on the drive as input. This seems possible, but need a 2nd opinion.

Many thanks
 

There's a thread about Fnirsi 138 Pro on EEVBlog .

Some quotes: "They're horrible." "The real bandwith is 50-100kHz".

If you want an ok/good 'scope at a reasonable price, look at e.g. Rigol, Siglent, Micsig.

There's a very important note in the AMC app note linked by Todd Zuercher (page 3):

Note: If a current probe is not available, use the drive current monitor pin. The signal from this pin is unfiltered and may be difficult to view. Also depending on the drive, this pin may not be isolated from the drive power ground. If this is the case, the oscilloscope must have isolated channels to avoid large ground currents.

Essentially this means that if you don't use a current probe (e.g. Micsig CP2100A, €180+VAT), you absolutely have to figure out if your drive's current monitor pin is of the non-isolated type. If it is, you need either a differential probe (these start at approximately €150+VAT - e.g. Micsig DP10013) or a (handheld) scope with isolated channels (these are quite expensive).

arvidb, Many, many thanks!! I am confident the approach I am working will work out, it is just going to take me some time to try the options until a good tune is found, I am taking good pics with descriptive file names so I can review what the trace looked like, if I need to review.

arvidb, thank you!

The data sheet doesn't exactly say if the current monitor is isolated. But it does say that it's output is 2A/V measured relative to the signal ground. Which for the drive's max current rating of 15A would only be 30v. Which sounds like an isolated signal to me since it is measured relative to the "signal ground". but maybe I'm wrong.